reporter‐based fate mapping in human kidney organoids ... · the capacity to create a model of...
TRANSCRIPT
Article
Reporter-based fate mapping in human kidneyorganoids confirms nephron lineage relationshipsand reveals synchronous nephron formationSara E Howden1,2,* , Jessica M Vanslambrouck1,2, Sean B Wilson1, Ker Sin Tan1 &
Melissa H Little1,2,3,**
Abstract
Nephron formation continues throughout kidney morphogenesis inboth mice and humans. Lineage tracing studies in mice identified aself-renewing Six2-expressing nephron progenitor population ableto give rise to the full complement of nephrons throughout kidneymorphogenesis. To investigate the origin of nephrons within humanpluripotent stem cell-derived kidney organoids, we performed asimilar fate-mapping analysis of the SIX2-expressing lineage ininduced pluripotent stem cell (iPSC)-derived kidney organoids toexplore the feasibility of investigating lineage relationships in dif-ferentiating iPSCs in vitro. Using CRISPR/Cas9 gene-edited lineagereporter lines, we show that SIX2-expressing cells give rise tonephron epithelial cell types but not to presumptive uretericepithelium. The use of an inducible (CreERT2) line revealed a declin-ing capacity for SIX2+ cells to contribute to nephron formation overtime, but retention of nephron-forming capacity if provided anexogenous WNT signal. Hence, while human iPSC-derived kidneytissue appears to maintain lineage relationships previously identi-fied in developing mouse kidney, unlike the developing kidneyin vivo, kidney organoids lack a nephron progenitor niche capableof both self-renewal and ongoing nephrogenesis.
Keywords CRSIPR/Cas9; fate mapping; kidney organoid; lineage tracing; pluri-
potent stem cell
Subject Categories Development & Differentiation; Stem Cells
DOI 10.15252/embr.201847483 | Received 27 November 2018 | Revised 5
February 2019 | Accepted 8 February 2019
EMBO Reports (2019) e47483
Introduction
Our understanding of kidney morphogenesis is predominantly based
on studies performed in model organisms. Gene-knockout and fate-
mapping studies, performed largely in mice, have uncovered many
of the fundamental molecular processes underlying kidney devel-
opment, homeostasis, and disease [1]. While inferences can be
made using a well-characterized mammalian model, our capacity
to validate the true relevance of lineage relationships or key
genetic pathways in human development is hampered by the scar-
city of human fetal tissue for experimental purposes. Recent analy-
ses of human fetal kidney have provided important insights and
highlighted several differences between humans and mice at
both the immunohistochemical and transcriptional levels [2–4].
However, the examination of humans fetal tissue is ethically
constrained, provides only snapshots at a fixed developmental
time-point, and does not represent an ideal platform for evaluating
whether differences between human and model organisms convey
functional relevance.
The capacity to create a model of the developing human kidney
in vitro may provide a unique opportunity to better understand
nephrogenesis at the molecular level in a human context. This
will depend, however, on the accuracy with which these models
of human kidney organogenesis recapitulate normal development.
Several protocols have now been described for the directed
differentiation of human pluripotent stem cells (hPSCs) to kidney
organoids [5]. We have established a protocol that generates
complex multicellular kidney organoids containing patterning and
segmenting nephrons and endothelial, perivascular, and stromal
cells [6,7]. After day 25 of differentiation, a single organoid
contains up to 100 nephrons, each beginning to show functional
maturation with podocyte foot process formation and albumin
uptake in proximal tubules. Moreover, kidney organoids exhibit a
transcriptional profile that is remarkably similar to first-trimester
human fetal kidney [7]. Combined with an ability to generate
developing kidney tissue in vitro, improvements in the speed and
accuracy with which it is possible to edit the genome of hPSCs
using CRISPR/Cas9 technology [8] provide a unique opportunity
to interrogate the molecular and cellular basis of morphogenesis
within kidney organoids to better understand these model
systems.
1 Murdoch Children’s Research Institute, Parkville, Vic., Australia2 Department of Paediatrics, The University of Melbourne, Melbourne, Vic., Australia3 Department of Anatomy and Neuroscience, The University of Melbourne, Melbourne, Vic., Australia
*Corresponding author. Tel: +614 73 019 385; E-mail: [email protected]**Corresponding author. Tel: +613 993 66 206; E-mail: [email protected]
ª 2019 The Authors EMBO reports e47483 | 2019 1 of 13
During kidney morphogenesis in the mouse, new nephrons form
throughout development from a mesenchymal population located at
the periphery of the developing kidney [9,10]. Marked by expression
of Six2 and Cited1, these cells exist in close association with the tips
of the branching ureteric epithelium. Nephron formation involves a
mesenchymal-to-epithelial transition to form a renal vesicle with
this process being accompanied by downregulation of Six2 [11,12].
In a seminal fate-mapping study performed in mice, Kobayashi et al
demonstrated that Six2-expressing cells represent a multipotent self-
renewing nephron progenitor population that, after undergoing
mesenchymal-to-epithelial transition, gives rise to all cells of the
murine nephron, but not to those within the branching ureteric
epithelium [9,10], which is instead derived from a distinct ureteric
progenitor population [13].
In this study, we chose to perform a similar fate-mapping analy-
sis of the SIX2-expressing lineage in iPSC-derived kidney organoids.
Using iPSCs harboring Cre recombinase within the endogenous SIX2
locus, in addition to a ubiquitously expressed loxP-flanked fluores-
cence cassette, we demonstrate that SIX2-expressing cells can
indeed contribute to nephron formation. While SIX2-derived cells
formed proximal nephron segments, they were absent from the
GATA3+CDH1+ epithelium, consistent with a ureteric epithelium
identity for the latter. Importantly, the use of an inducible CreERT2
lineage reporter showed a loss of contribution to new nephrons
across time within organoids, highlighting an absence of a nephro-
genic zone in such models. Despite this, SIX2+ cells could contri-
bute to new nephrons late in organoid culture if provided an
exogenous WNT signal, suggesting that they retain a nephron
progenitor identity across time even in the absence of a ureteric tip.
These findings not only improve our understanding of nephrogene-
sis within kidney organoids, but provide a proof of concept that
organoid and gene-editing technologies can be combined to interro-
gate and dissect human lineage relationships in vitro.
Results
SIX2 expression persists throughout kidneyorganoid differentiation
We first generated a SIX2 fluorescent reporter iPSC line to evaluate
and monitor SIX2-expressing cells within differentiating kidney
organoids in real time. Here, CRISPR/Cas9 was used to knock-in
EGFP at the 30 end of the SIX2 coding region, linked by the T2A self-
cleaving peptide (Fig 1A). Several clonally derived iPSC lines with
either heterozygous or homozygous insertion of the EGFP reporter
were established using a previously described method that combi-
nes reprogramming and gene editing together in a single step [8,14].
Both heterozygous (SIX2EGFP/+) and homozygous (SIX2EGFP/EGFP)
clones, distinguished by PCR analysis (Fig EV1A), were used in
subsequent differentiation experiments. Kidney organoids were
generated using our previously described protocol [6] (Fig 1B) with
two modifications: (i) TeSR-E6 was used instead of APEL as the base
differentiation medium, and (ii) a 3D bioprinter was used to gener-
ate day 7 aggregates via extrusion bioprinting onto Transwell filters
instead of manual transfer with a handheld pipette. This modified
protocol enables the generation of large numbers of highly repro-
ducible organoids that are equivalent at the level of morphology,
component cell types, and gene expression to those previously
reported via manual generation [preprint: 15]. Reporter gene expres-
sion in SIX2EGFP organoids was monitored routinely by flow cytome-
try and fluorescent microscopy and was first detected at
approximately days 8–10 of differentiation, consistent with previous
reports [13,16,17]. Notably, reporter gene expression was steadily
maintained after this time-point, until the cessation of our differenti-
ation experiments at day 25 (Fig 1C and D). Co-localization of EGFP
and SIX2 was confirmed by immunofluorescence (Fig 1E), with
EGFP- and SIX2-expressing cells restricted largely to the interstitial/
mesenchymal compartment (Fig EV1B), as anticipated. Co-localiza-
tion of EGFP and SIX2 was also confirmed by RT–PCR analysis of
sorted EGFP-expressing and non-expressing fractions (Fig EV1C).
Organoids derived from either SIX2EGFP/+ or SIX2EGFP/EGFP iPSCs
exhibited highly similar dynamics with respect to reporter gene
expression and nephron formation, although reporter gene intensity
was greater in SIX2EGFP/EGFP kidney organoids (Fig EV1D). EGFP
expression was also dependent on CHIR99021 concentration and
duration during the first stage of differentiation (Fig EV1E), consis-
tent with previous studies showing that increased WNT signaling
during hPSC differentiation induces a more posterior intermediate
mesoderm [7]. Organoids generated from cultures treated with
< 4 lM CHIR99021 failed to induce SIX2 expression or any recog-
nizable epithelial structures, whereas differentiations performed
with 8 lM CHIR for ≥ 4 days contained the highest fraction of SIX2-
expressing cells (> 50%). Cultures treated with 6–8 lM CHIR99021
for 4 days formed organoids with the most balanced profile with
respect to relative abundance of NEPHRIN+ podocytes, LTL+ proxi-
mal tubules, ECAD+ distal tubule structures, and putative
GATA3+/CDH1+ collecting duct epithelium, as determined by
whole-mount immunostaining (Fig 1F). This condition was used for
all subsequent differentiations.
Single-cell RNAseq of kidney organoids reveals widespreadSIX2-expression
To examine SIX2-expressing cells in greater depth during the
process of kidney organoid differentiation, single-cell transcriptome
profiling of day 18 and 25 organoids was performed using the 10×
Chromium platform. Twenty cell populations emerged from guided
clustering analyses using Seurat [18], with several of these pertain-
ing to different nephron segments and cells at various stages of
nephrogenesis (Fig 2A, Dataset EV1). Multiple stromal populations
were also identified, which expressed collagens COL1A1 and
COL3A1, as well as kidney stromal markers DCN and CXCL12. We
note the apparent absence of an endothelial cluster in this dataset,
despite clear evidence for this cell type in previous studies [19–21],
including in bioprinted organoids [preprint: 15].
With respect to SIX2-expressing cells, a distinct population (clus-
ter 9) exhibited strong congruence with human fetal nephron
progenitors, as determined by co-expression of several other previ-
ously described nephron progenitor markers, including CITED1,
DAPL1, EYA1, and SALL1 [3,13,16] (Fig 2B). Notably, SIX2 expres-
sion was not solely restricted to the putative nephron progenitor
cluster, with SIX2 transcripts detected in scattered cells within
several additional clusters, including a subset of the renal stroma,
and within an “off-target” population that displayed a muscle-like
transcriptional profile (Fig 2C). Interestingly, this SIX2-expressing
2 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
muscle-like population was clearly evident in day 25 organoids, but
largely absent in earlier day 18 organoids (Fig 2C and D). Compared
to day 18, day 25 organoids also showed an overall reduction in
both the SIX2-expressing nephron progenitor and “early committed
nephron” clusters (Fig 2D). Taken together, these findings reveal
that SIX2 is not confined to the presumptive nephron progenitor
compartment within human kidney organoids, but also marks
several additional distinct cell types, including renal stroma and a
muscle-like population which becomes more prevalent as the dif-
ferentiation proceeds.
Generation of a dual-fluorescence cassette for human fate-mapping studies
To facilitate human fate-mapping experiments in iPSC-derived orga-
noids, we generated a dual-fluorescence gene-targeting cassette
comprising a loxP-flanked EGFP and adjacent mCherry reporter that
can be incorporated into the 30 end of the endogenous GAPDH
coding region (Fig 3A). We have previously shown that this target-
ing strategy facilitates ubiquitous and consistent transgene expres-
sion in hPSCs, both prior to and following differentiation into
various different cell types [22]. We first evaluated the functionality
of our fluorescence cassette in the human embryonic stem cell line,
H9. Following CRISPR/Cas9-mediated knock-in of our targeting
cassette, correctly edited cells were identified based on expression
of the EGFP reporter (Fig 3A). To validate the Cre-mediated color
switching capacity of our dual-fluorescence cassette, EGFP-expres-
sing clones were isolated, expanded, and subsequently transfected
with mRNA encoding Cre recombinase. This resulted in the rapid
induction of mCherry expression and a corresponding loss of EGFP
expression within 8 h post-transfection in > 95% of cells (Fig 3B).
Kidney organoids could also be successfully derived from EGFP or
ex 1 SIX2ex 2
EGFP pAT2AA
E
C
CHIR (6 µM) FGF9 (200 ng/µl) No GFs (E6 medium)
0
iPSCs
AggregationCHIR pulse
2D culture (LN521) 3D culture (transwell)
4 7 12 25Day
B
SIX2 EGFP Merge
D0 D7 D13 D19 D250
10203040506070
% E
GFP
+ ce
lls
Day of differentiation
F EPCAM LTL NPHS1 GATA3Merge
D
50
100
150
200
250
0 310 410 5100 310 410 510
Day 0
EGFP= 0.3%
50
100
150
200
250
0 310 410 510
Day 10
EGFP= 48.6%
Day 25
0 310 410 510
50
100
150
200
250
EGFP= 53.0%
EGFP
SSC
Figure 1. Analysis of SIX2-expressing cells during kidney organoid development.
A Schematic diagram of the targeting strategy for generation of SIX2 reporter iPSCs. The EGFP gene was inserted just upstream of the SIX2 stop codon, linked via theself-cleaving T2A sequence.
B Outline of the kidney differentiation protocol used throughout this study.C Flow cytometry analysis of kidney organoids derived from SIX2EGFP/EGFP iPSCs.D Flow cytometry analysis of SIX2EGFP/EGFP kidney organoids shows EGFP/SIX2 expression emerges after day 7 of differentiation and persists thereafter. Data represent
mean � SD, n = 3.E Immunostaining of a day 12 SIX2EGFP/EGFP kidney organoid confirming co-localization of SIX2 and EGFP.F Immunofluorescence of SIX2EGFP/EGFP organoid demonstrating expression and correct localization of nephron segment-specific markers for proximal tubules
(EPCAM+/LTL+; blue/green), distal tubules (EPCAM+; green), collecting duct (EPCAM+/GATA3+; green/red), and glomeruli (NPHS1+; gray).
Data information: Scale bars = 50 lm.
ª 2019 The Authors EMBO reports e47483 | 2019 3 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports
mCherry-expressing H9 cells, representing cells before and after
exposure to Cre recombinase, respectively. We observed mainte-
nance of appropriate reporter gene expression in all component cell
types as determined by live fluorescent microscopy and flow cytom-
etry (Fig 3C and D).
Tracing the fate of SIX2+ cells in human kidney organoids
To examine the fate of SIX2-expressing cells in hPSC-derived kidney
organoids, we used CRISPR/Cas9-mediated gene editing to insert
the Cre recombinase gene into the endogenous SIX2 locus using the
targeting strategy described earlier for generation of SIX2EGFP iPSCs.
Clonally derived iPSCs with homozygous insertion of Cre recombi-
nase were established using our one-step reprogramming/gene-
editing protocol [8] which were subsequently used for knock-in of
the dual-fluorescence cassette into the GAPDH locus as described
above (Fig 4A). Correctly targeted EGFP-expressing colonies, here-
after referred to as SIX2Cre/Cre:GAPDHdual iPSCs, were identified by
fluorescent microscopy, isolated, and expanded for downstream dif-
ferentiation experiments. Kidney organoids were generated from
SIX2Cre/Cre:GAPDHdual iPSCs and monitored by flow cytometry for
reporter gene expression. Cells expressing mCherry could be
detected at approximately day 10 of differentiation, coinciding with
activation of endogenous SIX2 and reporter expression within
kidney organoids derived from the SIX2EGFP iPSCs described earlier
(Figs 4B and 1C). As differentiation progressed, a steady increase in
mCherry-expressing cells and a corresponding loss of EGFP-expres-
sing cells were also observed (Fig 4B). Live mCherry+/EGFP� cells
could also be detected in kidney organoids by fluorescent micro-
scopy, some of which appeared to be localized within epithelial
structures (Fig 4C). Whole-mount immunofluorescence of day 25
SIX2Cre/Cre:GAPDHdual organoids was performed to determine the
precise location of mCherry-expressing cells within specific cellular
compartments, using markers specific to nephrons (WT1, NPHS1,
LTL, CDH1, EPCAM), presumptive ureteric epithelium (GATA3,
CDH1), renal interstitium (MEIS1), and endothelium (CD31). SIX2-
expressing cells were seen to contribute to nephron formation, as
evidenced by the appearance of mCherry+ cells within LTL+ proxi-
mal tubules, EPCAM+/LTL� distal tubules, and NPHS1+ podocytes
(Fig 4D). Consistent with our scRNAseq analysis, interstitial cells
co-expressing MEIS1 and mCherry were also clearly evident, indicat-
ing that SIX2+ cells give rise to at least a subset of the renal stroma
(Fig 4D). Conversely, CD31+/mCherry+ cells were not observed,
indicating that endothelial cells within kidney organoids are not
A B
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
20
Cluster
Per
cent
age
of o
rigin
alda
tase
t in
each
clu
ster Day 25
Day 18
Day 18 Day 25SIX2
0 Max
15
10
5
0
CITED1 DAPL1
EYA1 SALL10 Max
0123456789
101112131415161718
C D
19
19
PodocytesNPHS1NPHS2MAFB
Proximal TubuleMT1GHNF4ACUBNLRP2
Distal Tubule & Loop of Henle
SLC12A1POU3F3
IRX1
Collecting DuctGATA3
TMEM213Epithelium (cycling)CCND1EPCAM
MuscleMYOD1MYOG
StromaCRABP1CRABP2
SIX1
StromaCRABP1CRABP2
SIX1
Nephron ProgenitorsSIX2
CITED1DAPL1LYPD1
Committed & Early Nephron
LYPD1LHX1JAG1
NeuralNTRK2
PTNFABP7
Cell CycleCCNB1CCNB2CENPECENPF
StromaStromaCommitted & Early NephronStromaStromaStromaProximal TubulePodocytesStromaNephron ProgenitorsDistal Tubule & Loop of HenleMuscleNeuralStromaPodocytesEpithelium (Cell Cycle)Collecting DuctCell CycleMuscleCommitted and Early Nephron
Figure 2. Analysis of SIX2 expression in kidney organoids by single-cell RNAseq.
A tSNE plot representing the merged datasets from single-cell RNAseq analysis of day 18 (1,865 cells) and day 25 (3,500 cells) organoids. Twenty clusters were identified.B A distinct SIX2-expressing cluster was identified as a nephron progenitor-like population based on expression of other known markers, including CITED1, DAPL1, EYA1,
and SALL1.C SIX2 marks a diverse range of cell types within human kidney organoids with SIX2 transcripts detected in numerous clusters in both day 18 and day 25 organoids.D Graphical representation showing the cell populations that are enriched or depleted in day 18 versus day 25 kidney organoids.
4 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
derived from SIX2+ cells (Fig 4D). Notably, mCherry+ cells were
excluded from GATA3+/CDH1+ structures (Fig 4E), which is
consistent with a ureteric epithelium-like identity for this population
of cells, as previous fate-mapping analyses in mice demonstrate
SIX2+ cells do not give rise to the ureteric epithelium [10]. Using
image analysis software, we detected < 1% of mCherry+ cells
within GATA3+/CDH1+ epithelium, with most of these events
observed at the proximal end of GATA3+/CDH1+ structures
(Fig 4F). Taken together, our findings are consistent with previous
studies performed in mice, which show SIX2+ cells can give rise to
all cells of the nephron but not the collecting duct network, which is
instead derived from a more anterior ureteric progenitor population
[10,13] (Fig 4G).
A SIX2-expressing population gives rise to nephrons early inorganoid culture but retains nephron-forming capacity acrossorganoid differentiation
In the developing kidney in vivo, new nephrons arise throughout
fetal development, arising from a self-renewing Six2+ nephron
progenitor population present around the tips of the ureteric
epithelium [10]. This process continues until approximately week
36 in humans [23,24] and the first few days after birth in mice
[25], at which point all NPCs have committed to nephron forma-
tion. To determine the duration of nephrogenesis across the
period of kidney organoid culture, we generated SIX2 knock-in
iPSCs using the tamoxifen-inducible Cre recombinase, CreERT2
[26]. The dual-fluorescence cassette was subsequently inserted
into the endogenous GAPDH locus of a clonally derived iPSC line
harboring a homozygous insertion of CreERT2 (Fig 5A). Day 12
kidney organoids derived from SIX2CreERT2/CreERT2:GAPDHDual
iPSCs were cultured in the presence of 4-hydroxytamoxifen (4-
OHT) for 1 h to induce Cre recombination. Activation of the
mCherry reporter could be detected by flow cytometry and fluo-
rescent microscopy within 24 h post-treatment (Fig 5B). A dose-
dependent trend was also noted, with the number of mCherry+
cells positively correlating with 4-OHT concentration as antici-
pated (Fig 5C). Importantly, no mCherry+ cells were observed in
the absence of 4-OHT treatment.
To determine if SIX2+ cells could contribute to nephron forma-
tion throughout organoid development, we staggered the labeling of
SIX2-expressing cells by initiating 4-OHT treatment (1 lM) at 2-day
intervals between days 10 and 18 of differentiation (Fig 6A). Orga-
noids were dissociated at day 25, stained with a directly conjugated
EPCAM antibody, and analyzed by flow cytometry to determine the
percentage of mCherry+ cells localized within epithelial structures.
ex 1
EGFP mCherrypAT2A
GAPDHex 9
pA
loxP loxP
A B
GAPDHdual iPSCs
Phase EGFP mCherry
GAPDHdual iPSCs+Cre mRNA
ex 1
mCherryT2A
GAPDHex 9
pA
loxP
+Cre
0 103
10 4 10 50 103
10 4 10 5
-Cre+Cre
-Cre +Cre
EGFP
coun
t
mCherry
C D
+ C
re
- Cre
EGFP mCherry
Brightfield EGFP
mCherry
mCherry
Merge EGFP mCherry Merge
Figure 3. Generation and characterization of a dual fluorescent reporter construct for downstream lineage tracing experiments.
A Schematic diagram of the targeting strategy. A loxP-flanked EGFP and adjacent mCherry reporter were inserted downstream of the endogenous GAPDH codingregion, linked via a self-cleaving T2A sequence. In the absence of Cre recombinase, cells constitutively express EGFP.
B The loxP-flanked EGFP gene is deleted following Cre-mediated recombination. Shortly after (< 8 h) the transient transfection of mRNA encoding Cre recombinase,GAPDHdual hPSCs permanently switch from EGFP to mCherry reporter gene expression.
C, D Reporter gene expression is maintained in all cell types within kidney organoids generated from GAPDHdual hPSCs before and after exposure to Cre, as detected byfluorescent microscopy (C) and flow cytometry (D).
Data information: Scale bar = 50 lm (panels A and B) and 200 lm (panel C).
ª 2019 The Authors EMBO reports e47483 | 2019 5 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports
ex 1 SIX2ex 2
Cre pAT2A
ex 1
EGFP mCherrypAT2A
GAPDHex 9
pA
loxP loxP
Proximal tubules & NPC Stroma
ME
IS1/
2/3
NP
HS
1m
Che
rry
Vasculature
CD
31LT
Lm
Che
rry
mC
herr
y
Glomeruli
EP
CA
M N
PH
S L
TLm
Che
rry
NP
HS
LT
Lm
Che
rry
EP
CA
MS
IX1
LTL
mC
herr
y
NP
HS
1m
Che
rry
EP
CA
M
SIX
1m
Che
rry
Distal tubule
EP
CA
MLT
Lm
Che
rry
mC
herr
y
mC
herr
y G
ATA
3m
Che
rry
EC
AD
mCherry
GAT
A3+/
ECAD
+
E
LHX1OSR1PAX2
SIX2CITED1
WT1
HOXB7RETPAX2
Nephron progenitors
Ureteric epithelium
Collecting duct
Nephrons
F G
mCherry+ 55.96%
Day 14
mCherry+ 13.37%
Day 10
mCherry+ 33.23%
Day 12
mCherry+ 0.28%
69% 060Day 7
mCherryE
GFP
+
A B C
D
Brightfield mCherry
EGFP + mCherryEGFP
LTL
Figure 4. Fate mapping of the SIX2 population in hPSC-derived kidney organoids.
A Schematic diagram of the targeting strategy used for generation of SIX2Cre/Cre:GAPDHdual iPSCs.B Flow cytometry analysis of kidney organoids derived from SIX2Cre/Cre:GAPDHdual iPSCs showing induction of mCherry and corresponding loss of EGFP expression.C Low (upper panel)- and high-magnification (lower panel) images showing mCherry+ cells detected by live fluorescent microscopy in SIX2Cre/Cre:GAPDHdual kidney organoids.D Immunostaining of SIX2Cre/Cre:GAPDHdual kidney organoids shows localization of mCherry cells within proximal (LTL+/EPCAM+), distal (LTL�/EPCAM+), and glomerular
(NPHS1+) nephron segments and within renal stroma (MEIS1+) but not within the CD31+ vasculature.E mCherry+ cells were excluded from the presumptive GATA3+/ECAD+ collecting duct epithelium.F Plot showing exclusion of mCherry+ cells from GATA3+/ECAD+ epithelium as determined by image analysis software.G Model depicting the separation of nephron and collecting duct lineages during kidney morphogenesis.
Data information: Scale bars = 50 lm.
6 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
Using this assay, we observed significantly fewer EPCAM+ cells
within the mCherry+ fraction of organoids induced at day 18
compared to those induced at day 10 (Figs 6B and EV2A). While a
negative correlation between the time of 4-OHT treatment and
EPCAM+/mCherry+ cells was observed, there was no correlation
between time of 4-OHT treatment and the total number of
mCherry+ cells (Fig EV2B) nor the overall fraction of EPCAM+ cells
(Fig EV2C). Whole-mount immunofluorescence of day 25
SIX2CreERT2/CreERT2:GAPDHdual organoids was also performed, where
we observed mCherry+ cells both within EPCAM+ structures and
the interstitial compartment, but not within the GATA3+/EPCAM+
epithelium (Fig 6C). In organoids induced early, at day 10,
mCherry+ cells were easily detected in all nephron epithelial
segments and EPCAM�/NEPHRIN+ podocytes (Fig 6D, early induc-
tion). Conversely, in organoids induced at later time-points,
mCherry+ cells were notably absent within nephron structures and
predominantly restricted to the interstitium surrounding them
(Fig 6D, late induction). Because our transcriptional profiling data
revealed a SIX2-expressing cell population with strong similarity to
human nephron progenitors in day 18 kidney organoids (see Fig 2),
we considered whether an additional CHIR pulse could promote
SIX2+ cells in late (day 18) kidney organoids to undergo nephron
commitment. In this set of experiments, kidney organoids induced
with 4-OHT at day 18 were subsequently subjected to a 1-h CHIR
pulse the following day. When the organoids were harvested 1 week
later (day 26) and analyzed by whole-mount immunofluorescence,
we could clearly detect mCherry+ cells derived from the SIX2+
lineage in EPCAM+ epithelial structures and NEPHRIN+ podocytes,
whereas mCherry+ cells were restricted to the interstitium in control
organoids that had not undergone a late CHIR pulse. Collectively,
these findings indicate that the capacity for SIX2+ cells to contribute
to nephron formation is retained across organoid culture, but
requires an exogenous differentiation signal such as can be provided
via induction of WNT signaling.
Discussion
Organoids derived from hPSCs offer enormous utility in personal-
ized disease modeling and drug testing platforms, while also provid-
ing promise for the development of autologous cellular therapies to
treat/correct many inherited and acquired diseases. Organoid-based
cultures also represent a potential source of human tissue at devel-
opmental stages that are typically unavailable for research purposes.
In combination with gene-editing technologies, this could facilitate
the study of gene function and cellular processes that govern human
development in vitro. Genome engineering also facilitates studies
aimed at characterizing the cell types produced in organoid-based
cultures and how well these compare with primary developing
tissue, which is imperative for understanding and addressing limita-
tions associated with hPSC-derived tissue.
In this study, we use gene-edited iPSCs to interrogate the SIX2
lineage in human kidney organoids and to examine how this
compares with our existing understanding of mammalian nephroge-
nesis in vivo, which is largely based on studies performed in mice.
We first used our previously described kidney organoid differentia-
tion protocol and a SIX2EGFP reporter line to monitor SIX2+ cells in
developing kidney organoids. SIX2+ cells emerged soon after orga-
noid formation and persisted until the termination of differentiation
at day 25. Although this contrasts to other previously described
ex 1 SIX2ex 2
CreERT2 pAT2A
ex 1
EGFP mCherrypAT2A
GAPDHex 9
pA
loxP loxP+
A
B CEGFP + mCherry
00.1 0.5 1 2 10
1234567
0.05
4-OHT concentration (µM)%
mC
herr
y+ c
ells
Figure 5. Temporal and limited labeling of the SIX2 lineage using the inducible Cre recombinase, CreERT2.
A Schematic diagram of the targeting strategy used for generation of SIX2CreERT2/CreERT2:GAPDHdual iPSCs.B mCherry+ cells detected by live fluorescent microscopy in SIX2CreERT2/CreERT2:GAPDHdual kidney organoid following 4-OHT induction at day 10 of differentiation.
Epithelial mCherry+ cells are indicated (white arrow). Scale bar = 50 lm.C The number of mCherry+ cells in kidney SIX2CreERT2/CreERT2:GAPDHdual organoids positively correlates with 4-OHT concentration. Data represent mean � SD, n = 3.
ª 2019 The Authors EMBO reports e47483 | 2019 7 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports
differentiation protocols where a rapid loss of SIX2+ cells is
observed soon after the formation of 3D kidney organoids [16], our
findings are consistent with a recent study that employed a similar
SIX2-EGFP reporter iPSC line and kidney organoid protocol, where
SIX2 expression also persisted until the termination of differentia-
tion [27]. Importantly, our findings do not support their hypothesis
that the presence of these cells definitively represents a progenitor
niche that may supply the organoid with more mature cells over
time. Rather, we show that SIX2 is expressed in a variety of other
cell types, and a lack of ongoing nephrogenesis across organoid
culture. While single-cell transcriptome profiling of organoids gener-
ated in this study did reveal a distinct SIX2+ population that exhib-
ited strong congruence with human fetal nephron progenitors, SIX2
expression was also detected in several additional cell clusters.
Interestingly, this included an “off-target” muscle-like population
that was enriched in late but not early organoids. SIX2 expression
was also detected in several clusters identified as “renal stroma”.
Consistent with this observation, our fate-mapping experiments
revealed an obvious SIX2+ lineage-derived MEIS1+ stromal popula-
tion. Although this contrasts sharply with studies performed in
mice, where a strict lineage boundary between nephron progenitors
and the interstitial progenitor cells that give rise to the renal stroma
has been shown to exist [28–30], recent single-cell RNAseq analysis
of human fetal kidney has revealed substantial overlap between
these two progenitor populations, with co-expression of SIX2,
MEIS1, and FOXD1 detected within human nephron progenitors [3].
Although difficult to determine definitively whether the co-expres-
sion of nephron progenitor and stromal markers is an artifact of our
organoid culture system, the findings from this previous study
suggest that this may in fact be a true species difference. In mice,
the origin of the endothelial population has not been regarded to be
the SIX2-expressing population and we did not see evidence for
B
10 12 14 16 18Day of 4-OHT induction
%E
PC
AM
+ m
Che
rry
cells
18
14
10
6
2
DE
arly
indu
ct.
Late
indu
ct.
* p=0.0027
EPCAM GATA3 DAPI mCherry EPCAM LTL NPHS1 mCherry
CHIR(6 µM)
FGF9 (200 ng/ µl) No GFs
0 4 7 10 12 14 16 18 25
2D (LN521) 3D (Transwell)
Day
4-OHT initiation
SIX2
A
C
EPC
AM
LTL
NPH
S1m
Che
rry
EPC
AM
GA
TA3
DA
PIm
Che
rry
+CH
IR D
ay 1
9C
ontro
l
E +4-OHT Day 18
Figure 6. Staggered labeling of SIX2 lineage shows declining nephrogenic potential of SIX2+ cells during kidney organoid differentiation.
A Outline of the strategy used to interrogate the nephrogenic potential of SIX2+ cells throughout kidney organoid differentiation.B The percentage of mCherry+ cells localized within EPCAM+ epithelial structures negatively correlates with the time of 4-OHT initiation, as determined by flow
cytometric analysis. Data represent mean � SD, n ≥ 3. Data from day 10 and 18 time-points were obtained from two independent experiments. P-value wascalculated using unpaired two-sided t-test.
C Immunostaining of SIX2CreERT2/CreERT2:GAPDHdual kidney organoids shows mCherry+ cells localized within EPCAM+ structures (white arrow) and the interstitialcompartment (yellow arrow), but not within the GATA3+/EPCAM+ epithelium.
D mCherry+ cells derived from the SIX2+ lineage were detected in all nephron epithelial segments when 4-OHT induction was initiated at day 10 (early induction) butwere largely restricted to the interstitium when induced at later time-points (late induction).
E In late kidney organoids induced with 4-OHT at day 18, mCherry+ cells derived from the SIX2+ lineage were detected in nephron segments when subjected to a late(day 19) CHIR pulse. mCherry+ cells were restricted to the interstitium in control organoids that had not undergone a late CHIR pulse.
Data information: Scale bars = 50 lm.
8 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
endothelial cells of SIX2 lineage in this study; however, endothelial
cell types were low in abundance, and hence, the lineage relation-
ship here remains equivocal and requires more investigation.
Further studies are also warranted to determine which sub-popula-
tions of the overall SIX2-expressing population can in fact contribute
to nephron formation in our organoid model.
The presence of cell clusters with a muscle signature that appar-
ently increase in prevalence with time is difficult to interpret. We
would note, however, the prevalent formation of ectopic muscle
within Wilms’ tumors, a childhood renal neoplasia regarded as show-
ing disrupted development [31]. It has previously been suggested
that Wilms’ tumor arises as a result of inappropriate differentiation of
the nephron progenitors [31,32], with this conclusion recently
supported at the single-cell level [33]. It has also been suggested that
muscle formation within Wilms’ tumors is associated with reductions
in WT1 expression. Hence, it is possible that this off-target popula-
tion is arising from the initial nephron-forming SIX2 population.
Using a lineage tracing iPSC reporter line, we demonstrate that
SIX2-expressing cells can contribute to all segments of the develop-
ing nephron but are excluded from the distal ends of the GATA3+/
CDH1+ epithelial structures, suggestive of a presumptive ureteric
epithelium population in kidney organoids. Of note, not all cells
within each nephron had undergone Cre-mediated color switching,
with many nephrons comprised of both mCherry+ and EGFP+ cells.
One explanation for this is that Cre recombination may not have
been complete. Alternatively, there may be a genuine contribution
of both SIX2-expressing and non-expressing cells within forming
nephrons. More studies are required to investigate this further.
By using a tamoxifen-inducible variant of Cre recombinase to
stagger the labeling of SIX2+ cells during organoid differentiation,
we noted a significantly reduced capacity for these cells to contri-
bute to nephron formation over time. This suggests human kidney
organoids, in contrast to fetal kidney in vivo, lack a true nephro-
genic zone capable of sustained nephrogenesis. In mice, correct
localization of nephron and stromal progenitors around the tips of
the ureteric epithelium enables reciprocal inductive signals between
these populations. This is required for continued branching of the
ureteric epithelium and both nephron progenitor self-renewal and
commitment, which drives organogenesis throughout fetal develop-
ment [12,34]. Interestingly, although SIX2+ cells in more mature
(day 18) organoids did not contribute to nephron formation, these
cells could in fact give rise to nephrons upon the addition of a late
(day 19) CHIR pulse. This finding is consistent with the transcrip-
tional profiling of day 18 organoids, where a SIX2-expressing popu-
lation with strong congruence with human nephron progenitors was
identified. However, it suggests that this competent population
ceases nephron formation in the absence of an exogenous differenti-
ation signal. Hence, although we demonstrate clear evidence of
nephron progenitor commitment in our organoid cultures, a self-
renewing niche comprised of a branching ureteric epithelium with
corresponding tips and a distinct domain of nephron and stromal
progenitors is not evident. This suggests that appropriate spatial
organization and/or reciprocal interactions between nephron
progenitors, ureteric epithelium, and possibly also the renal stroma
are deficient.
Perhaps a more logical strategy for generating kidney organoids
with a sustainable nephrogenic niche would be to derive homoge-
nous populations of nephron and stromal progenitors, and ureteric
epithelium in parallel which could be subsequently combined in 3D
culture, in a spatial arrangement that more accurately depicts the
developing organ in vivo. This strategy was partially recapitulated in
a recent study which demonstrated a substantially improved higher-
order structure in kidney organoids derived from mouse PSCs, and
was highlighted by an impressive capacity for the ureteric epithe-
lium to undergo several rounds of branching morphogenesis [35].
However, while a human branching ureteric epithelium was gener-
ated, there was not a successful reciprocal interaction between this
epithelium and surrounding presumptive human metanephric
mesenchyme, highlighting our limited understanding of the exact
conditions required to recapitulate a self-renewing nephron progeni-
tor niche coupled with ongoing productive nephron formation
in vitro.
In conclusion, our findings show that nephrons within kidney
organoids arise from a SIX2-expressing mesenchymal population, as
anticipated from previous studies in mice, but also show the
absence of ongoing nephrogenesis, likely due to the lack of a periph-
eral nephrogenic zone. However, at least a subset of SIX2-expressing
cells retain nephron-forming capacity longer term and can form
nephrons if induced to do so. As such, this represents a proof of
concept that gene-editing and organoid technologies can be
combined to facilitate fate-mapping studies in differentiating hPSCs,
thereby providing a unique opportunity to investigate lineage rela-
tionships in real time and in a higher-throughput and more cost-
effective manner compared with mammalian models. Additional
fate-mapping and CRISPR/Cas9-mediated gene-knockout studies in
organoids may facilitate the development of more efficient and
robust protocols to generate renal cell types for downstream applica-
tions. Indeed, such approaches may also provide deeper insight into
human kidney development. While applied to kidney in this
instance, such a lineage tracing approach is applicable in other orga-
noid settings in which complex multicellular tissues are formed.
Materials and Methods
Cell lines
Human foreskin fibroblasts (ATCC ID: CRL-2429) were cultured in
DMEM (Thermo Fisher Scientific) supplemented with 15% fetal
bovine serum (FBS; Hyclone) and 1X MEM Non-Essential Amino
Acids Solution (Thermo Fisher Scientific) at 37°C, 5% CO2, and 5%
O2. All iPSC lines were maintained and expanded at 37°C, 5% CO2,
and 5% O2 in Essential 8 medium (Thermo Fisher Scientific) on
Matrigel-coated plates with daily medium changes and passaged
every 3–4 days with EDTA in 1X PBS as previously described [36].
The genomic integrity of iPSCs was confirmed by molecular kary-
otyping using Infinium CoreExome-24 v1.1 SNP arrays (Illumina),
and expression of common pluripotency markers (TRA-1-81, SSEA-
4, CD9, OCT4) was confirmed by immunofluorescence and flow
cytometry.
Kidney organoid production
The day prior to differentiation, cells were dissociated with TrypLE
(Thermo Fisher Scientific), counted using a hemocytometer, and
seeded onto Laminin 521-coated 6-well plates at a density of
ª 2019 The Authors EMBO reports e47483 | 2019 9 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports
50 × 103 cells per well in Essential 8 medium. Intermediate meso-
derm induction was performed by culturing iPSCs in TeSR-E6
medium (Stem Cell Technologies) containing 4–8 lM CHIR99021
(R&D Systems) for 4 days. On day 4, cells were switched to TeSR-E6
medium supplemented with 200 ng/ml FGF9 (R&D Systems) and
1 lg/ml Heparin (Sigma-Aldrich). On day 7, cells were dissociated
with TrypLE, diluted fivefold with TeSR-E6 medium, transferred to
a 15-ml conical tube, and centrifuged for 5 min at 300 × g to pellet
cells. The supernatant was discarded, and cells were resuspended in
residual medium and transferred directly into a syringe for bioprint-
ing. Syringes containing the cell paste were loaded onto a NovoGen
MMX Bioprinter, primed to ensure cell material was flowing, and
user-defined aliquots (5,000–100,000 cells per organoid) were
deposited on 0.4-lm Transwell polyester membranes in 6-well
plates (Corning). Following bioprinting, organoids were cultured for
1 h in the presence of 6 lM CHIR99021 in TeSR-E6 medium in the
basolateral compartment and subsequently cultured until day 12 in
TeSR-E6 medium supplemented with 200 ng/ml FGF9 and 1 lg/ml
Heparin. From day 12 to day 25, organoids were grown in TeSR-E6
medium without supplementation. Unless otherwise stated, kidney
organoids were cultured until harvest at day 25. For induction of
CreERT2 protein in kidney organoids derived from SIX2CreERT2
iPSCs, 4-hydroxytamoxifen (Sigma-Aldrich) dissolved in ethanol at
a concentration of 100 lM was diluted to working concentration in
TeSR-E6. Induction medium (1 ml) was pipetted under the tran-
swell, and individual drops from a 20-ll pipette were carefully
placed on top of the organoids on the filter to ensure complete
coverage. Organoids were incubated at 37°C for 1 h. Induction
medium was removed by washing three times with TeSR-E6
medium every 10 min and then returned to the media they were in
prior to induction.
Vector construction
The SIX2:EGFP vector (pDNR-SIX2:EGFP) carries a targeting cassette
encoding the T2A peptide and EGFP gene flanked by ~700 and
~450 bp of sequence corresponding to sequence immediately
upstream and downstream of the SIX2 stop codon, respectively.
Two gBlocks (Integrated DNA Technologies) encoding the targeting
cassette were inserted into the AatII and EcoRI sites of the pDNR-
Dual (Clontech) plasmid vector. The pDNR-SIX2:Cre and pDNR-
SIX2:CreERT2 targeting vectors were generated as described above
but substituting the Cre recombinase and CreERT2 recombinase
genes for EGFP, respectively. The GAPDH targeting vector encoding
the dual-fluorescence cassette (pGAPTrap-loxEGFPloxCherry) was
generated by inserting sequence encoding the T2A peptide, loxP-
flanked EGFP gene with SV40 polyA signal and adjacent mCherry
gene with SV40 polyA signal into the SfiI and ClaI sites of the pGAP-
Trap-mtagBFP2-IRESMuro plasmid vector (after removal of the
mtagBFP2-IRESMuro cassette). A sgRNA plasmid specific to the 30
end of the SIX2 coding region (pSMART-sgRNA-SIX2) was generated
by annealing ODNs SIX2_sgRNA1a and SIX2_sgRNA1b followed by
ligation into the BbsI sites of the pSMART-sgRNA vector [37]. A
sgRNA plasmid specific to the 30 end of the GAPDH coding region
(pSMART-sgRNA-GAPDH) was generated by annealing ODNs
GAPDH_sgRNA1a and GAPDH_sgRNA1b followed by ligation into
the BbsI sites of the pSMART-sgRNA vector. All plasmids were prop-
agated in DH5-alpha Escherichia coli (BIOLINE) and prepared for
transfection using a Plasmid Maxi kit (QIAGEN). See Table 1 for list
of ODNs and Table 2 for list of plasmids used in this study.
Generation of knock-in iPSCs
All SIX2 knock-in iPSCs (EGFP, Cre, CreERT2) were derived from
human foreskin fibroblasts (ATCC: CRL-2429) using a previously
described protocol that combines reprogramming and gene editing in
one step [8]. Episomal reprogramming plasmids (pEP4E02SET2K,
pEP4E02SEN2L, pEP4E02SEM2K, and pSimple-miR302/367), in vitro
transcribed mRNA encoding the SpCas9-Gem variant [37], the
pSMART-sgRNA-SIX2 plasmid, and either the pDNR-SIX2:EGFP,
pDNR-SIX2:Cre, or pDNR-SIX2:CreERT2 targeting vectors were intro-
duced into fibroblast using the Neon Transfection System as
described below. In vitro transcribed mRNA encoding a truncated
version of the EBNA1 protein was also included to enhance nuclear
uptake of the reprogramming plasmids [36,38]. Genomic DNA was
isolated from resulting iPSCs using the DNeasy Blood & Tissue Kit
(QIAGEN) in accordance with the manufacturer’s protocol, and PCR
analysis was performed using GoTaq Green Master Mix (Promega)
Table 1. Oligonucleotides used in this study.
Name Sequence
GAPDH_sgRNA1a CACCGCTTCCTCTTGTGCTCTTGCT
GAPDH_sgRNA1b AAACAGCAAGAGCACAAGAGGAAGC
SIX2_sgRNA1a CACCGTCAGCCAACCTCGTGGACC
SIX2_sgRNA1b AAACGGTCCACGAGGTTGGCTGAC
SIX2F CATCTACCCAGCAAACCTGG
EGFPR GTCCAGCTCGACCAGGATGG
SV40pAF GCGACTCTAGATCATAATC
SIX2R GAGTACAAGAGACTGGCAGG
CreR GAGTTGATAGCTGGCTGGTG
Table 2. Plasmids used in this study.
Name References Identifier
pEP4 E02S ET2K Yu et al [39] Addgene plasmid #20927
pEP4 E02S EN2L Yu et al [39] Addgene plasmid #20922
pEP4 E02S EM2K Yu et al [39] Addgene plasmid #20923
pSimple-miR302/367 Howden et al [14] Addgene plasmid #98748
pSP6-EBNA2A+DBD Howden et al [38] Addgene plasmid #98749
pDNR-SpCas9-Gem Howden et al [14] Addgene plasmid #98749
pSMART-sgRNA(Sp) Howden et al [14] Addgene plasmid #80427
pGAPTrap-mtagBFP2-IRESMuro
Kao et al [22] Addgene plasmid #82335
pGAPTrap-loxEGFPloxCherry
This study TBC
pDNR-Dual Clontech N/A (discontinued)
pDNR-SIX2:EGFP This study TBC
pDNR-SIX2:Cre This study TBC
pDNR-SIX2:CreERT2 This study TBC
10 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
to identify correctly targeted clones. ODNs SIX2F and EGFPR flank the
50 recombination junction of SIX2:EGFP knock-in iPSCs, whereas
SIX2F and CreR flank the 50 recombination junction of SIX2:Cre and
SIX2:CreERT2 knock-ins. ODNs SV40pAF and SIX2R flank the 30
recombination junction of SIX2:EGFP, SIX2:Cre, and SIX2:CreERT2
knock-in iPSCs. Heterozygous and homozygous clones were
distinguished using ODNs SIX2F and SIX2R. For knock-in of the dual-
fluorescence cassette, the GAPDH targeting construct (pGAPTrap-
loxEGFPloxCherry) was co-transfected with pSMART-sgRNA-GAPDH
and mRNA encoding SpCas9-Gem into hPSCs using the Neon
Transfection System as described below. Correctly targeted (EGFP-
expressing) clones were identified by fluorescent microscopy.
In vitro transcription
Capped and polyadenylated in vitro transcribed mRNA encoding
SpCas9-Gem protein [37] was generated using the mMESSAGE
mMACHINE T7 ULTRA transcription kit (Thermo Fisher Scientific)
according to the manufacturer’s recommendations. Plasmid
template was linearized with PmeI endonuclease prior to transcrip-
tion. LiCl was used to precipitate mRNA before resuspension. A
truncated version of the EBNA1 protein [38], used to facilitate
uptake of reprogramming plasmids, was transcribed using the
mMESSAGE mMACHINE SP6 transcription kit (Thermo Fisher
Scientific) according to the manufacturer’s recommendations.
Cell transfection
Transfections were performed using the Neon Transfection System
(Thermo Fisher Scientific). Human fibroblasts or iPSCs were
harvested with TrypLE (Thermo Fisher) 2 days after passaging and
resuspended in Buffer R at a final concentration of 1 × 107 cells/ml.
Electroporation was performed in a 100-ll tip using 1,400 V, 20 ms,
and 2 pulses for human fibroblasts, or 1,100 V, 30 ms, and 1 pulse for
human iPSCs. Following electroporation fibroblasts were transferred
to 6-well Matrigel-coated plates containing DMEM + 15% FBS and
switched to reprogramming medium (TeSR-E7 + 100 lM sodium
butyrate) after 3 days, with medium changes every other day. Electro-
porated human iPSCs were plated on 6-well Matrigel-coated plates
containing Essential 8 medium with 5 lM Y-27632 (Tocris).
Flow cytometry
Prior to analysis, single kidney organoids were dissociated with
0.2 ml of a 1:1 TrypLE/Accutase solution in 1.5-ml tubes at 37°C for
15–25 min, with occasional mixing (flicking) until large clumps
were no longer clearly visible. 1 ml of HBBS supplemented with 2%
FBS was added to the cells before passing through a 40-lM FACS
tube cell strainer (Falcon). Flow cytometry was performed using a
LSRFortessa Cell Analyzer (BD Biosciences). Data acquisition and
analysis were performed using FACSDiva (BD) and FlowLogic soft-
ware (Inivai). Gating was performed on live cells based on forward-
and side-scatter analysis.
Whole-mount immunofluorescence
Organoids were transferred to 48-well plates for fixation and
immunofluorescence procedures. Fixation was performed using
ice-cold 2% paraformaldehyde (PFA; Sigma-Aldrich) for 20 min
followed by 15 min of washing in three changes of phosphate-
buffered saline (PBS). For immunofluorescence, blocking and anti-
body staining incubations were performed on a rocking platform for
3 h at room temperature or overnight at 4°C, respectively. Blocking
solution consisted of 10% donkey serum with 0.3% Triton X-100
(TX-100; Sigma-Aldrich) in PBS. See Table 3 for list of antibodies
used in this study. Antibodies were diluted in 0.3% TX-100/PBS.
Primary antibodies were detected with Alexa Fluor-conjugated fluo-
rescent secondary antibodies (Invitrogen), diluted 1:500. Organoids
were washed in at least three changes of PBS for a minimum of 1 h
following primary and secondary antibody incubations. Imaging
was performed in glass-bottomed dishes (MatTek) with glycerol
submersion using either the Zeiss LSM 780 or Dragonfly Spinning
Disk confocal microscope.
Single-cell transcriptional profiling and data analysis
Organoids were dissociated as described above (for flow cytometry)
and passed through a 40-lM FACS tube cell strainer. Following
centrifugation at 300 g for 3 min, the supernatant was discarded
and cells resuspended in 50 ll TeSR-E6 medium. Viability and cell
number were assessed, and samples were run across separate runs
on a Chromium Chip Kit (10× Genomics). Libraries were prepared
using Chromium Single Cell Library kit V2 (10× Genomics) and
Table 3. Antibodies used in this study.
Name Source Identifier
Mouse monoclonalanti-TRA-1-81
BioLegend Cat#330706
Mouse monoclonal anti-SSEA-4 BioLegend Cat#330408
Mouse monoclonal anti-CD9 BD Biosciences Cat#555371
Rabbit monoclonal anti-OCT4 Abcam Cat#Ab181557
Rabbit polyclonalanti-RFP (and mCherry)
MBL Medical &BiologicalLaboratories
Cat#PM005
Biotinylated LotusTetragonolobus Lectin
Vector Laboratories Cat#B-1325
Rabbit polyclonal anti-SIX1 Cell Signaling Cat#12891
Rabbit polyclonal anti-SIX2 Proteintech Group Cat#11562-1-AP
Mouse monoclonalanti-MEIS1/2/3
Active Motif Cat#39795
Mouse anti-EPCAM(Alexa Fluor 488-conjugated)
BioLegend Cat#324210
Mouse anti-EPCAM(Alexa Fluor 647-conjugated)
BioLegend Cat#324212
Mouse monoclonal anti-GATA3 Thermo FisherScientific
Cat#MA1-028
Goat polyclonal anti-GATA3 R&D Systems Cat#AF2605
Sheep polyclonal anti-NEPHRIN R&D Systems Cat#AF4269
Mouse monoclonal anti-CD31 BD Pharmingen Cat#550274
Mouse monoclonalanti-E-CADHERIN
BD Biosciences Cat#610181
Chicken polyclonal anti-GFP Abcam Cat#Ab13970
ª 2019 The Authors EMBO reports e47483 | 2019 11 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports
sequenced on an Illumina HiSeq with 100-bp paired-end reads. Cell
Ranger (v1.3.1) was used to process and aggregate raw data from
each of the samples returning a count matrix. Quality control and
analysis was performed in R using the Seurat package (v2.3.1) [18].
Cells with more than 125,000 UMIs, < 500 genes expressed, or more
than 20% reads assigned to mitochondrial genes were filtered out.
UMI counts, percentage of mitochondrial and ribosomal gene
expression, and cell cycle phase identity were regressed out. Genes
with less than two counts across the whole dataset were also filtered
out. The final dataset had 5,365 cells and 22,105 identified genes.
The two samples were merged using a canonical correlation analysis
(CCA) using 1,429 genes with the highest dispersion present in both
samples. The CCA subspaces were aligned, and the first 25 principal
components based on these genes were used to build a graph, which
was clustered at a resolution of 1.6. Data from the single-cell tran-
scriptional profiling have been deposited in the Gene Expression
Omnibus under accession number GSE119561.
Expanded View for this article is available online.
AcknowledgementsThe Murdoch Children’s Research Institute is supported by the Victorian
Government’s Operational Infrastructure Support Program. The MCRI Gene
Editing Facility is supported by the Stafford Fox Foundation. MHL is a
Senior Principal Research Fellow of the National Health and Medical
Research Council, Australia (GNT1136085). This work was supported by the
National Institutes of Health, USA (DK107344-01), and the NHMRC
(GNT1100970).
Author contributionsSEH and MHL conceived the study and wrote the manuscript; SEH performed
gene-editing experiments; SEH, JMV, SBW, and KST performed kidney
differentiation experiments; JMV, SBW, and KST performed immunostaining;
SBW performed scRNAseq analysis; and all authors assisted in manuscript
preparation.
Conflict of interestM.H.L. is an inventor on a patent associated with kidney organoid generation
and has a research contract with and has consulted for Organovo Inc.
References
1. Humphreys BD, DiRocco DP (2014) Lineage-tracing methods and the
kidney. Kidney Int 86: 481 – 488
2. O’Brien LL, Guo Q, Lee Y, Tran T, Benazet JD, Whitney PH, Valouev A,
McMahon AP (2016) Differential regulation of mouse and human
nephron progenitors by the Six family of transcriptional regulators.
Development 143: 595 – 608
3. Lindstrom NO, Guo J, Kim AD, Tran T, Guo Q, De Sena Brandine G,
Ransick A, Parvez RK, Thornton ME, Basking L et al (2018) Conserved
and divergent features of mesenchymal progenitor cell types within the
cortical nephrogenic niche of the human and mouse kidney. J Am Soc
Nephrol 29: 806 – 824
4. Lindstrom NO, McMahon JA, Guo J, Tran T, Guo Q, Rutledge E, Parvez
RK, Saribekyan G, Schuler RE, Liao C et al (2018) Conserved and diver-
gent features of human and mouse kidney organogenesis. J Am Soc
Nephrol 29: 785 – 805
5. Morizane R, Miyoshi T, Bonventre JV (2017) Concise Review: kidney
generation with human pluripotent stem cells. Stem Cells 35:
2209 – 2217
6. Takasato M, Er PX, Chiu HS, Little MH (2016) Generation of kidney orga-
noids from human pluripotent stem cells. Nat Protoc 11: 1681 – 1692
7. Takasato M, Er PX, Chiu HS, Maier B, Baillie GJ, Ferguson C, Parton RG,
Wolvetang EJ, Roost MS, Chuva de Sousa Lopes SM et al (2015) Kidney
organoids from human iPS cells contain multiple lineages and model
human nephrogenesis. Nature 526: 564 – 568
8. Howden SE, Thomson JA, Little MH (2018) Simultaneous reprogramming
and gene editing of human fibroblasts. Nat Protoc 13: 875 – 898
9. Boyle S, Misfeldt A, Chandler KJ, Deal KK, Southard-Smith EM, Mortlock
DP, Baldwin HS, de Caestecker M (2008) Fate mapping using Cited1-
CreERT2 mice demonstrates that the cap mesenchyme contains self-
renewing progenitor cells and gives rise exclusively to nephronic epithe-
lia. Dev Biol 313: 234 – 245
10. Kobayashi A, Valerius MT, Mugford JW, Carroll TJ, Self M, Oliver G,
McMahon AP (2008) Six2 defines and regulates a multipotent self-
renewing nephron progenitor population throughout mammalian
kidney development. Cell Stem Cell 3: 169 – 181
11. Short KM, Combes AN, Lefevre J, Ju AL, Georgas KM, Lamberton T, Cairn-
cross O, Rumballe BA, McMahon AP, Hamilton NA et al (2014) Global
quantification of tissue dynamics in the developing mouse kidney. Dev
Cell 29: 188 – 202
12. Little MH, McMahon AP (2012) Mammalian kidney development: princi-
ples, progress, and projections. Cold Spring Harb Perspect Biol 4:
a008300
13. Taguchi A, Kaku Y, Ohmori T, Sharmin S, Ogawa M, Sasaki H, Nishinaka-
mura R (2014) Redefining the in vivo origin of metanephric nephron
progenitors enables generation of complex kidney structures from
pluripotent stem cells. Cell Stem Cell 14: 53 – 67
14. Howden SE, Maufort JP, Duffin BM, Elefanty AG, Stanley EG, Thomson JA
(2015) Simultaneous reprogramming and gene correction of patient
fibroblasts. Stem Cell Reports 5: 1109 – 1118
15. Higgins JW, Chambon A, Bishard K, Hartung A, Arndt D, Brugnano J,
Xuan Er P, Lawlor KT, Vanslambrouck JM, Wilson S et al (2018)
Bioprinted pluripotent stem cell-derived kidney organoids provide
opportunities for high content screening. bioRxiv https://doi.org/10.1101/
505396505396 [PREPRINT]
16. Morizane R, Lam AQ, Freedman BS, Kishi S, Valerius MT, Bonventre
JV (2015) Nephron organoids derived from human pluripotent stem
cells model kidney development and injury. Nat Biotechnol 33:
1193 – 1200
17. Takasato M, Er PX, Chiu HS, Maier B, Baillie GJ, Ferguson C, Parton RG,
Wolvetang EJ, Roost MS, Lopes SM et al (2016) Kidney organoids from
human iPS cells contain multiple lineages and model human nephroge-
nesis. Nature 536: 238
18. Butler A, Hoffman P, Smibert P, Papalexi E, Satija R (2018) Integrating
single-cell transcriptomic data across different conditions, technologies,
and species. Nat Biotechnol 36: 411 – 420
19. Phipson B, Er PX, Combes AN, Forbes TA, Howden SE, Zappia L, Yen HJ,
Lawlor KT, Hale LJ, Sun J et al (2019) Evaluation of variability in human
kidney organoids. Nat Methods 16: 79 – 87
20. Combes AN, Zappia L, Er PX, Oshlack A, Little MH (2019) Single-cell
analysis reveals congruence between kidney organoids and human fetal
kidney. Genome Med 11: 3
21. Wu H, Uchimura K, Donnelly EL, Kirita Y, Morris SA, Humphreys BD
(2018) Comparative analysis and refinement of human PSC-derived
12 of 13 EMBO reports e47483 | 2019 ª 2019 The Authors
EMBO reports Fate mapping in human kidney organoids Sara E Howden et al
kidney organoid differentiation with single-cell transcriptomics. Cell
Stem Cell 23: 869 – 881 e868
22. Kao T, Labonne T, Niclis JC, Chaurasia R, Lokmic Z, Qian E, Bruveris FF,
Howden SE, Motazedian A, Schiesser JV et al (2016) GAPTrap: a simple
expression system for pluripotent stem cells and their derivatives. Stem
Cell Reports 7: 518 – 526
23. Hinchliffe SA, Sargent PH, Howard CV, Chan YF, van Velzen D (1991)
Human intrauterine renal growth expressed in absolute number of
glomeruli assessed by the disector method and Cavalieri principle. Lab
Invest 64: 777 – 784
24. Ryan D, Sutherland MR, Flores TJ, Kent AL, Dahlstrom JE, Puelles VG,
Bertram JF, McMahon AP, Little MH, Moore L et al (2018) Development
of the human fetal kidney from mid to late gestation in male and
female infants. EBioMedicine 27: 275 – 283
25. Rumballe BA, Georgas KM, Combes AN, Ju AL, Gilbert T, Little MH (2011)
Nephron formation adopts a novel spatial topology at cessation of
nephrogenesis. Dev Biol 360: 110 – 122
26. Indra AK, Warot X, Brocard J, Bornert JM, Xiao JH, Chambon P, Metzger
D (1999) Temporally-controlled site-specific mutagenesis in the basal
layer of the epidermis: comparison of the recombinase activity of the
tamoxifen-inducible Cre-ER(T) and Cre-ER(T2) recombinases. Nucleic
Acids Res 27: 4324 – 4327
27. Borestrom C, Jonebring A, Guo J, Palmgren H, Cederblad L, Forslow A,
Svensson A, Soderberg M, Reznichenko A, Nystrom J et al (2018) A CRISP
(e)R view on kidney organoids allows generation of an induced pluripo-
tent stem cell-derived kidney model for drug discovery. Kidney Int 94:
1099 – 1110
28. Kobayashi A, Mugford Joshua W, Krautzberger AM, Naiman N, Liao J,
McMahon Andrew P (2014) Identification of a multipotent self-renewing
stromal progenitor population during mammalian kidney organogenesis.
Stem Cell Reports 3: 650 – 662
29. Mugford JW, Sipilä P, McMahon JA, McMahon AP (2008) Osr1 expression
demarcates a multi-potent population of intermediate mesoderm that
undergoes progressive restriction to an Osr1-dependent nephron
progenitor compartment within the mammalian kidney. Dev Biol 324:
88 – 98
30. Levinson R, Mendelsohn C (2003) Stromal progenitors are important for
patterning epithelial and mesenchymal cell types in the embryonic
kidney. Semin Cell Dev Biol 14: 225 – 231
31. Miyagawa K, Kent J, Schedl A, van Heyningen V, Hastie ND (1994) Wilms’
tumour–a case of disrupted development. J Cell Sci Suppl 18: 1 – 5
32. Schedl A (2007) Renal abnormalities and their developmental origin. Nat
Rev Genet 8: 791 – 802
33. Young MD, Mitchell TJ, Vieira Braga FA, Tran MGB, Stewart BJ, Ferdinand
JR, Collord G, Botting RA, Popescu DM, Loudon KW et al (2018) Single-
cell transcriptomes from human kidneys reveal the cellular identity of
renal tumors. Science 361: 594 – 599
34. Combes AN, Lefevre JG, Wilson S, Hamilton NA, Little MH (2016) Cap
mesenchyme cell swarming during kidney development is influenced by
attraction, repulsion, and adhesion to the ureteric tip. Dev Biol 418:
297 – 306
35. Taguchi A, Nishinakamura R (2017) Higher-order kidney organogenesis
from pluripotent stem cells. Cell Stem Cell 21: 730 – 746 e736
36. Chen G, Gulbranson DR, Hou Z, Bolin JM, Ruotti V, Probasco MD,
Smuga-Otto K, Howden SE, Diol NR, Propson NE et al (2011) Chemically
defined conditions for human iPSC derivation and culture. Nat Methods
8: 424 – 429
37. Howden SE, McColl B, Glaser A, Vadolas J, Petrou S, Little MH, Elefanty
AG, Stanley EG (2016) A Cas9 variant for efficient generation of Indel-
free knockin or gene-corrected human pluripotent stem cells. Stem Cell
Reports 7: 508 – 517
38. Howden SE, Wardan H, Voullaire L, McLenachan S, Williamson R, Ioan-
nou P, Vadolas J (2006) Chromatin-binding regions of EBNA1 protein
facilitate the enhanced transfection of Epstein-Barr virus-based vectors.
Hum Gene Ther 17: 833 – 844
39. Yu J, Hu K, Smuga-Otto K, Tian S, Stewart R, Slukvin II, Thomson JA
(2009) Human induced pluripotent stem cells free of vector and trans-
gene sequences. Science 324: 797 – 801
ª 2019 The Authors EMBO reports e47483 | 2019 13 of 13
Sara E Howden et al Fate mapping in human kidney organoids EMBO reports